Keynotes | K&Ts | GACs | Talks | Posters | Search
Poster C in Poster Session C: Wednesday, August 5, 9:30 – 11:15 am, Kimmel Center, Shorin & Rosenthal Rooms
Grasp-relevant Representation of Objects in the Human Parietal and Occipito-temporal Cortices
Amirali Soltani Tehrani1, Kristin Woodard2, Emalie McMahon3, Leah Ettensohn4, Natalia Pallis-Hassani2, John Ingeholm5, Leslie Ungerleider2, Chris I. Baker5, Maryam Vaziri-Pashkam1; 1University of Delaware, 2National Institute of Mental Health, 3Massachusetts Institute of Technology, 4University of California, San Diego, 5National Institutes of Health
Presenter: Amirali Soltani Tehrani
The human visual system supports a broad range of behaviors toward objects, from recognizing them to performing actions on them. In this fMRI study, we focused on two specific behaviors (odd-one-out similarity judgments and grasping) and investigated how representations in the parietal and occipitotemporal cortices relate to each behavior. In the fMRI experiment, participants viewed and, after a delay, grasped 3D-printed objects. We localized object-selective regions in lateral occipitotemporal cortex (LOT) and inferior intraparietal sulcus (inferior IPS), as well as grasp-selective regions in anterior intraparietal area (AIP) and supramarginal gyrus (SMG). In the visual phase, LOT and inferior IPS responses were more strongly associated with the odd-one-out than grasping behavior, whereas AIP and SMG did not show this pattern. During the motor phase, AIP and SMG patterns related to both the grasp and the odd-one-out behaviors, while LOT and inferior IPS remained associated with just the odd-one-out behavior. All regions were associated with both behaviors during the delay phase. These findings highlight the distinct roles of occipitotemporal and parietal regions in a natural grasping task: occipitotemporal regions reflect object similarity independent of action, whereas parietal regions reflect both object similarity and grasp-relevant features, simultaneously encoding multiple representations to support coordinated actions toward objects.
Topic Area: Computational Models of Vision & Visual Cortex